TOPIC:

I recently purchased a miniDSP DDRC22D. After taking measurements with UMIK 1 and create the filters I want to share my experience with forum members because it is not completely satisfactory.
The main shortcoming is bass reproduction.
The curve proposed by Dirac makes the bass disappear completely. To solve this I am forced to manually modify the curve and increase +8 dB so that there is a presence of bass without being at all convincing, at least you can hear it.
Although I've never been a fan of exaggerated bass, initially I thought that this subjective absence of bass could be due to my personal preferences and listening habits. I brought a audiophile friend to give me their opinions and his veredict was: lack of bass.
Another effect I have noticed is an excess level at high frequencies. It is exactly the opposite of the bass: correction by Dirac left me with too much treble and therefore I am forced to manually correct to take it down to an acceptable point. Pre-equalization curve shows a sharp drop in level from 10 kHz and up which makes no sense to me since speakers and electronics have a bandwidth greater than 20 Hz - 20 kHz.
For me it is as if the microphone deceive Dirac indicating that there are more bass than there actually are and less high frequencies than there are in reality. That leads Dirac to reduce the bass and increase the treble.
I have verified that microphone calibration file correlates to serial number of the microphone.
What is going on here?
I'm sure I am doing something wrong, so any idea or advice will be welcome.

Just to add some perspective from a fellow Dirac user (DDRC-88A), my experiences are considerably different than yours. I measure the audio performance using the UMIK-1 microphone and REW, and I am seeing flat bass down to 20Hz, and a gentle roll-off in the high frequencies exactly as the Dirac Auto Target predicts. I mention this to illustrate that there is nothing inherent in Dirac correction that would result in the lack of bass and exaggerated high frequencies that you report. I would guess that you may have a procedural error in your calibration, or some other factor that is causing the disappointing results. Do you use REW and can you measure the Dirac results?

@ Juan,

We have quite a lot of these units now in the fields and your experience indeed sounds out of the norm. The DDRC-22D being a Digital in/out module, we basically don't have a way to modify the audio (i.e. it's all digital) "unless" Dirac somehow measures, assumes something else. So I think that the fact that you have issue at "both" the low and high, points to something else here.
a) What's your curve like in the measurement? i.e. maybe take a screenshot.
b) Did you perform all 8 measurements as advised in the user manual? Please clarify how these were done?
c) Did you insure that the microphone USB cable is correctly plugged in? i.e. hearing a click. Any extension cable used here? It almost sounds like it's a measurement issue at this point as already suggested by Jerry.

Keep us updated with more information, we'd most likely need to see your Dirac project to know more.

DevTeam

@Juan,

I'm also a DDRC-22D user and i'm very happy with it.
Could you take some screenshots of your measurements before and after targeting your curve?

Thanks to AustinJerry and Devteam for answering my post.

devteam wrote:

@ Juan,.
a) What's your curve like in the measurement? i.e. maybe take a screenshot.
b) Did you perform all 8 measurements as advised in the user manual? Please clarify how these were done?
c) Did you insure that the microphone USB cable is correctly plugged in? i.e. hearing a click. Any extension cable used here? It almost sounds like it's a measurement issue at this point as already suggested by Jerry.
DevTeam

a)I attached it
b) I did performed 8 measurements following user manual indications. Not once, but several times. I use chair configuration and try to separate the measurements 30 cm apart one from the next.
c) I am sure the microphone is correctly plugged in. Indeed, the first time I made a measurement I didn´t pluggen it in to the end and the mic wast not recognized.

AustinJerry wrote:

Do you use REW and can you measure the Dirac results?

I am attaching it.

By the way, I don´t doubt that Dirac and DDRC22D perform well. Before buying DDRC22D I made an extensive search on the internet, reading comments and reviews about it. I am sure too that there is something wrong on my side and I appreciate the help I am receiving to point the problem out.

Now I am attaching the screenshot of Dirac software.

@ Juan,

Ok, maybe the reason why you feel the DDRC-22D doesn't play bass is more a fact that by default (i.e. raw system), your current stereo setup actually does have a lot of bass. You can see upward of more than 10dB peaks compared to the base line/target curve. You understand what we mean?

So indeed, I'd expect a bit of a "shock" if you add a target curve with the autotarget that would be that far from how your system initially sounded (i..e lots of bass and roll off). That's expected because the autotarget (based on acoustic and making a balanced system) is just far from what you've been used to listen..

- Quick question to steer away from the Dirac correction for a sec: Have you ever measured your system (e.g. using REW) before this day? Roll off (even if speakers and equipment tell you they do 20-20k) isn't too unique. We've seen this before. The fact that bass of uncorrected system (light blue) is so hot and target curve (orange) is 10dB lower certainly explains why you'd feel it's lacking bass after Dirac..

- What's the screenshot of your system "after" correction. i.e. need to optimize and click on the "after" tick box at the top.

Defining the target curve is key element of any room correction software (Dirac or not). It takes a bit of time to figure out what you want to hear. Seeing the "after" plots and comparing with REW will tell you what was done and how it was done.

DevTeam

That dip at 35 Hz is huge and no EQ can fix that. Is there a wall behind the MLP? If so how far from the wall do you sit? Is your room a square or rectangle? What are the dimensions? Some fairly careful use of passive acoustic treatments can go a long way in evening out your frequency plot. That in turn will allow you to dial in a house curve to your liking. Combining Dirac with thoughtful use of passive acoustic treatments is the mother load!

Sent from my LG-VS980 using Tapatalk

@Juan,

I see that without Dirac correction you have a 20dB boost between 50 - 120Hz, could be roommode or if you use a seperate subwoofer which is bass heavy.
I think you have a roommode around 65Hz, if i ignore that i see that your bass is falling of around 55-60Hz

You use the default target curve (flat response) which indeed increases the highs.
It's better to adjust your target to follow your measured curve and do little tweaks.
See my curve and the target:



It's better to start with placement of speaker, acoustical treatment and then DSP.
Good luck.

devteam wrote:

@ Juan,

Ok, maybe the reason why you feel the DDRC-22D doesn't play bass is more a fact that by default (i.e. raw system), your current stereo setup actually does have a lot of bass. You can see upward of more than 10dB peaks compared to the base line/target curve. You understand what we mean?

So indeed, I'd expect a bit of a "shock" if you add a target curve with the autotarget that would be that far from how your system initially sounded (i..e lots of bass and roll off). That's expected because the autotarget (based on acoustic and making a balanced system) is just far from what you've been used to listen..

DevTeam

Your reasoning was my first line of thought. Honestly, I don´t think this is the case.
Please see my first post where I explain:

Juan wrote:

Although I've never been a fan of exaggerated bass, initially I thought that this subjective absence of bass could be due to my personal preferences and listening habits. I brought a audiophile friend to give me their opinions and his veredict was: lack of bass.
.

The absence of bass is evident not only for me, but also for other trained ears.

devteam wrote:

@ Juan,

- Quick question to steer away from the Dirac correction for a sec: Have you ever measured your system (e.g. using REW) before this day? Roll off (even if speakers and equipment tell you they do 20-20k) isn't too unique. We've seen this before. The fact that bass of uncorrected system (light blue) is so hot and target curve (orange) is 10dB lower certainly explains why you'd feel it's lacking bass after Dirac..

- What's the screenshot of your system "after" correction. i.e. need to optimize and click on the "after" tick box at the top.

DevTeam

I am attaching the screenshot of my system after correction.
I have measured my system using REW before and obviously the curve is similar to the red one I attached in my previous post.
I would like to pay attention to my previous comment about a possible problem with the microphone:

Juan wrote:

For me it is as if the microphone deceives Dirac indicating that there are more bass than there actually are and less high frequencies than there are in reality. That leads Dirac to reduce the bass and increase the treble.
I have verified that microphone calibration file correlates to serial number of the microphone.

If Dirac thinks that there is a lot of bass because the mic is giving this incorrected information it will reduce accordingly the bass level. The same would apply to the high frequencies.
Is this a possibilty or should I disregard it completely ?

jagman wrote:

That dip at 35 Hz is huge and no EQ can fix that. Is there a wall behind the MLP? If so how far from the wall do you sit? Is your room a square or rectangle? What are the dimensions? Some fairly careful use of passive acoustic treatments can go a long way in evening out your frequency plot. That in turn will allow you to dial in a house curve to your liking. Combining Dirac with thoughtful use of passive acoustic treatments is the mother load!

Sent from my LG-VS980 using Tapatalk

The wall behind the LP is at 2.20 meters ( 7.3 ft).
Dimensions of the room are: 4,65 m (front wall) by 5,40 m and height 2.47 m. ( 15.25 ft by 17.71 ft by 8,1 ft in height).
The use of passive acoustic treatment is discarded.
It´s not possible to change the possition of the speakers or the MLP.
This is why I looked at solutions like Dirac!

Juan,

I'm pretty sure there is a lost in translation being miles away and trying to explain each others on a forum.. So let us try again. Maybe others here can try to pitch in with other explanations.. We will highlight important key words here with " ". Make sure to pay attention so we're on the same page, hoping there isn't confusion.

a) There is "no" doubt that the Dirac processed audio will "lack" bass. I'm not sure how to word it better, but the way you have you target curve, the way you have your system measuring raw (both in REW & in Dirac Live), you "will" experience a taming of the bass with Dirac. You can clearly see a peak of upward 13dB or more at the low frequency. That's a "massive" peak. It's being lowered by Dirac autotarget to less than 13dB. That's a LOT of bass removed.

b) There is "no" doubt that the Dirac processed audio "will" sound brighter because right now, the unprocessed audio is naturally rolling off. You can see that the processed Dirac response is making this flat, once again "as per the target curve". It is once again the target curve that defines the audio response.

Now with these 2 statements hopefully out of the way, let's see how we can make it sound the way you like it.

1) There is a simple line of thought that isn't mentioned here in much of your emails and it's about "Tweaking your target curve". It's the "key" experience in this whole story. You believe it's Dirac or the microphone, we'd like to point you toward learning to tweak your target curve.
If you want your system to sound the way you're used to i.e. higher amount of bass, then simply get the target curve to "follow" the natural response of your speakers the way you're use to it. Do "not" use the auto target because that's based on an algorithm and study of psycho acoustic, that may just not fit your preferences. See the pic by Kingpin above. See how he's following the target?

2) The assumption that a microphone would boost 13dB of bass hasn't been seen to-date (and we've got lots of mics out there). It doesn't mean it's impossible, it just means that it would be a very very odd case.. Now at the high frequency, quick question: Did you by any chance make these measurements with the windscreen? (i.e. black little cover to the head?) If so, please remove it. That's "only" to be used when there is wind, which hopefully there won't be in your room.

Hoping this info helps.

DevTeam

Let's start in the beginning. Your attempt to achieve good sound is no different than anyone else's, and it begins with the measurement of the uncorrected audio, which you provided in an earlier post. It clearly shows a huge bass boost at 60Hz, and a fairly rapid roll-off in the high end. Dirac is just doing what you tell it to do when you select Auto Targets--it is smoothing the bass peak, and raising the high frequencies. If you are used to how the bass sounds with that big peak, then naturally the bass is going to sound weaker now. There is a time-honored saying in audio: flat bass sounds like no bass at all.

The first step towards improvement involves working on the un-corrected room response. The methods are well-documented. Move the sub(s) around until that peak is reduced. Add more subs if you can't find a sub placement that fixes the peak, or consider adding bass traps. The better the un-corrected room response is, the better job Dirac will do.

Once you have a reasonably flat Dirac result, that's the time to experiment with the target curve editor to adjust the response to your liking. You can add more bass, as well as add a high frequency roll-off, quite effectively using the target editor. Expecting Dirac to solve all your issues without giving it a good room to work with is unrealistic.

@Devteam
Thanks for your support.
You make clear that the possibility of Umik-1 doing strange things is remote so I will forget about it.
As you suggest I will focus on tweking the curve to my taste. I will mimic Kingpin curve. Thanks for the example.
@AustinJerry
This is a 2 channel stereo setup. No subs involved. No way of adding passive acoustic treatment to the room and no way of moving speakers or LP. Even on this scenario I think DDRC22D is doing a good job in reducing room modes acoustic effect, correcting an asymetry of the LP and improving spaciousness of the music.

Hello !

As said before, everything is normal.
A flat response (there is no flat response without acoustic treatment, but you understand me) is alway a choc the first time.

Jean Pierre Lafont a very famous ingenior said :
"An hifi system, it is an equipment with no calibration, in a bad place, with defects whose enjoy the owner".
(sorry for the traduction...)

You need to understand what is a "good/bad" sound.
And after you can try to have an balanced sound. And with Dirac, you can add just a little colour with curve to match your taste.

Careful, you have lot of bass, and again, it is very common in our room (often rectangular, etc.).
So don't believe that the original sound without Dirac is the good one !
But you can, of course, add bass if you want to with Dirac, again it is a very common setting.

As you can see in my screenshot, I have lot of bass. I love that, but I have too much !!
I keep the auto target curve to have a more flat response (increase high, balance the medium, etc.) and I push a little ONLY in the bass because I like that. But I keep the other setting !